More change in Hapi? – CometWatch 22 January

Today’s CometWatch mosaic comprises images taken by Rosetta’s navigation camera (NAVCAM) on 22 January, from a distance of 28.0 km from the centre of Comet 67P/Churyumov-Gerasimenko. The image resolution is 2.4 m/pixel and the individual 1024 x 1024 frames measure 2.4 km across. The mosaic measures 4 x 4 km.


Four image mosaic comprising images taken on 22 January 2015 by Rosetta’s Navigation Camera (NAVCAM). Rotation and translation of the comet during the imaging sequence make it difficult to create an accurate mosaic, and there may be some spurious spatial and intensity features as a result of the mosaic-making sequence, so always refer to the individual frames before performing any detailed comparison or drawing conclusions about any strange structures or low intensity extended emission. Credits: ESA/Rosetta/NAVCAM – CC BY-SA IGO 3.0

The mosaic looks down onto the Seth region on the larger lobe (left) and its transition into the smooth neck region of Hapi, with the smaller comet lobe to the right – the rim of Hatmehit just visible along the horizon.

One particularly interesting observation is that the curved, horseshoe-shaped markings in Hapi have appeared to change in appearance since the last published images of this region (for example, take a look at the CometWatch entries for 8 January and 9 December).

Of course, we cannot discount the role that the change in viewing geometry/illumination conditions play, but the markings that in today’s image can be seen about a third from the top of the lower left frame, seem a lot more developed than in the 8 January image, where there is perhaps only just a hint of them in the making.

Today’s image also affords a great view of the flat plateau with its thumbprint-like depressions along its edge. As discussed in previous entries, these features are considered possible precursors to fracture or collapse. (Click here for OSIRIS image of this feature.)

We look forward to seeing what other interesting features you spot in today’s image set!

The four frames and a montage of the 1024 x 1024 images are provided below:







  • logan says:

    Hi Emily. This zone is not evolving the way I expected. Is receiving more dust than emitting??

  • Robin Sherman says:

    Thanks Emily for another interesting image. As you point out the different viewpoints and illumination, can be deceptive. The horseshoe shapes are different, but I am still inclined to think they are the terrain beneath the gravel/dust layer showing through more as the top surface layer is lost or transported to different areas of Hapi. That depression bottom left in the 9 Dec image had several sharply defined features, but by January, they had been covered over and we could see that the surface material had flowed into the depression.

    Other things I spotted that seem to have changed, appear at the other end of Hapi Valley. The previous images showed one or two small bright patches, maybe 2-3 metres in size. They are still there, but further up the valley there are four or five new larger very bright patches, maybe 10-15 metres across. The assumption is these are areas where, either there are volatile ices just below the surface and the angle of illumination makes them visible, or the grave/dust has been removed and these are freshly exposed areas of volatile ices, quite possibly water ice, which we know to be more abundant in the Hapi region.

    The other interesting thing is on the far right of the image, the original site of Phillae’s touchdown in Agilkia can be see. Earlier images showed it as a slightly brighter patch, in this image it is quite a large very bright patch. Philae we know disturbed a lot of dust, probably exposing the sub-surface, or at least allowing more energy to reach the sub-surface and this has lead to increased activity/removal of dust, setting off a cycle of faster erosion than the surrounding areas.

    In the January 8 image, the surface of Hatmehit is showing increasing evidence of developing horseshoe features that are not evident on the OSIRIS search image, but their positions can be seen as colouration changes. Again is this a sign that the underlying topology is what we are seeing? If so how did those shapes form in the sub-surface?

    Lastly, top centre of Image A is a very large cave that has been seen before, but not from almost face on. There is just enough reflected light to see inside the cave. It would seem not to be the “lid” over a deep pit, the cave floor can be seen running, still almost level, to the back wall. The way the surface above the cave appears to have been pushed up in a curve suggests these caves/pits are the result of pressure “bubbles” that build up only a few metres below the surface. At some point one wall, or the roof fractures and a gaping hole is left. The weakest point is likely to be the side that gets the most sunlight, so within a local area all the caves/pits would tend to face the same direction. Just another of my crackpot theories. 😉

    • Daniel says:

      About the Hapi bright patches. Do you mean the large bright patches at roughly (in pixel coordinates with 0,0 in the upper left corner) 960x 80y in the A frame? You can see them already in a NAVCAM image from 24th September 2014 but also in the 9th December image that emily helpfully linked to. They might be brighter than before, but I’m not really that sure about that.
      I am inclined to agree that it is an active area either way. Emissions seems to come from the general area after all during other images. Might want to look at as an reminder/example along with the comments by Bill.

  • Daniel says:

    It seems we have a few additional horseshoes and that some of the existing ones moved a bit. Hapi is the most active area on the comet, but it is odd that we haven’t seen any of the same formations being created anywhere else so far. Unless they have come and gone elsewhere too, and we simply haven’t spotted them.

    I would like to ask others to comment on a potential change in the “Amphitheatre” in the Seth region. It seems like there could be at least one change there, but I’m torn on whether it is real or just an illusion. It might be real, but at the same time I could rationalize it simply as the light and shadows fooling me.

    The change is at a rock outcropping in the “amphitheatre” were there now seems to be a small depression next to it. I’ve included some cut outs of the affected area from a number of NAVCAM images, the top row is pre-change and the bottom row post-change.
    Credits: ESA/Rosetta/NAVCAM – CC BY-SA IGO 3.0
    (As an aside, do the blog keepers care if I ignore adding the credits in the future when posting modified ESA images on this blog?)
    NAVCAM images are left to right top to bottom from:
    24th September 2014, 26th September 2014, 2nd October 2014,
    7th December 2014, 8th January 2015, 22nd January 2015
    The images are from

    Another good image of the area in question is the 28th August OSIRIS image illustrating the active pit in Seth.

    • Robin Sherman says:

      Hi Daniel.

      It is difficult to say for sure, but the impression is that the outcrop is higher above the surface, in other words the level of the dust has decreased. The science team seem to believe that sublimating gases create sufficient “wind” to move dust on the surface. Scouring of dust in the lea of rock outcrops is a common phenomenon seen in desert environments on Earth. This area is also quite near to the large pit at the end of the “Amphitheatre”. Is there perhaps an upwelling of gas that overflows over the rim of the pit across the plain of the “Amphitheatre”, creating this apparent dust loss and the deepening scallop shapes at the cliff edge where the gas sinks over the edge down to the Hapi Valley below?

      I am increasingly inclined to think these horseshoe and semi-circular shaped depressions are reflections of the underlying solid surface below. Their shapes seem to change and some have been covered up again.

      The absorbance spectra from VIRTIS showed a significant shift to the blue in the Hapi region and one possible explanation was given as smaller particle sizes for the surface dust here. This may explain why we have not seen these changes elsewhere, the surface dust could be more readily transported by the venting gases in the Hapi region. The greater activity in this region may be the reason for the smaller particle size. We have seen how fragile the larger grains of comet material are, the slightest impact reducing them to piles of smaller grains a few micrometers in size, “interplanetary dust” effectively.

      Looking at the surface of Hatmehit, the only other place where glimpses of these semi-circular features have been seen, it appears to be a lot more gravel like, rather than mainly dust. I would add though, a comparable close up of the Hapi surface has not so far been released by the OSIRIS team, so this should be seen as being more in the realms of conjecture.

  • Marco says:

    I don’t see any major changes in Hapi, but I do notice a huge boulder that has been dislodged into the nearby Ampitheatre, from the cliff wall about 170 metres away.

    • Robin Sherman says:

      I agree that this looks like a large slab has moved, just like the one I highlighted in the OSIRIS Lander Search image. The OSIRIS team pointed to another in one of their papers, that they estimated to be 100 metres plus in size. None of them appear to be associated with steep slopes so the question is how do such huge slabs of material move such significant distances all in one piece?. The paper suggests it is due to sublimation pressure building up below the surface. I tend to think this might explain the semicircular caves and scarps we see, but these appear to be really good fits back into the original landscape. The pressure buildup would have then to be contained within very narrow fissures, rather than bubbles of expanding sublimation gases that seem to form the caves and pits.

      Would this process in part explain your fracture planes? Again is the amount of energy reaching this far below the surface viable? It seems to be, the science team think it is a credible scenario. With this initial possible fissure creation and expansion mechanism creating large scale fracture planes beneath large areas of surface, the energy from spin up could remove these large slabs from the comet entirely leaving behind the large flat fracture planes observed. I am thinking of those large flat surfaces on the ends of the big lobe in particular.

      As I said before, I was not sure that spin up effects could provide enough energy to remove such large slabs, but with this possible weakening/fracturing from an additional process, it would seem to be a lot more credible scenario, especially as we see most evidence for material being permanently lost at the extremities of the comet.

      • Marco says:

        I’ve had a good look at the fracture planes, and they really appear parallel rather than concentric at the scale of 67P. I’m sure that sublimation or thermal stress could loosen these planes, but there is no process that would make a flat boundary of anything at this scale. My conclusion is that 67P is a fragment of a much larger icy body, such as a centaur like Chiron. Chiron is found to have a diameter of around 200km, and is spinning with a period of less than 6 hours. Additionally it is found to have rings, which I can only presume is from shed material. Talking of it being a fragment of a larger body is probably a stalking horse for EU theories, but I really can’t think of what can form flat boundaries otherwise.

        • Judy Hawkins says:

          The thing about the EU theories is they heat up the pot, and there’s really nothing wrong with that, as long as I don’t feel like I have to argue with them. It motivates me to post comments that suggest that people read the science behind all this, and I’ve rarely been this motivated to post to an internet forum.

          I think 67P is a fragment of something larger (the Hyperion images are very suggestive to me!) that got broken up going around Jupiter at some point in the more or less distant past. I look at Hatmehit and it looks like the hole left after something sizeable — like, a lump roughly the size of the small lobe, or somewhat smaller — got pulled off it.

          Having gotten that idea in my head, now I’m seeing Imhotep as the scar where 67P as a whole separated from something larger.

          Yikes, I’m getting my own hobby horse going…

          • logan says:

            If 67P’s neck was symmetrical at forming, Then I would like to add a cutting plane [or conic segmentation] at former ‘night side’.

        • THOMAS says:

          @ Marco

          “Talking of it being a fragment of a larger body is probably a stalking horse for EU theories”.

          You have yourself already suggested the idea of 67P being “a fragment of a larger body” on another thread, and I indeed fully agree with you. But without the ice, of course….

      • THOMAS says:

        @ Robin Sherman

        “we see most evidence for material being permanently lost at the extremities of the comet.”

        Really? *What* evidence is there for this? I thought we all more or less agreed now that, on the contrary, the vast majority of the “material” has been, and clearly still is being, permanently lost from the NECK region.

        • Marco says:

          Robyn is talking about “missing slabs” leaving behind flat surfaces of Imhotep and Hatmehit. These are clearly not ablation features and neatly have at the surface under some dust and rocks flat planar *strata* that so elegantly point to the possibility of 67P being a fragment of a larger body. The edges of the flat strata are raised 50 or so metres indicating the possibility of a very thick slab being prised off from the comet forever, especially at the extremities. I can’t believe you haven’t jumped on this as further evidence in support of the concept of large fragments being separated from a parent body.

          • THOMAS says:

            Sorry, Marco, I honestly haven’t followed that particular discussion that closely.

            As regards the “concept of large fragments being separated from a parent body” if I understand your theory correctly, it’s a mere matter of centrifugal force in your book. I certainly haven’t “jumped on” it, because the mechanism proposed in the EU model is entirely different: asteroids and comets are the relics of untold billions of tons of material which were blasted off the surface of one of the rocky planets, in particular Mars, by catastrophic electric discharge activity.

          • Marco says:

            Hi THOMAS, leaving aside *mechanism* considerations, there can be evidence for something being in a particular spot and that spot being now vacant, without there being any direct evidence for the mechanism. If electricity is capable enough to excavate large chunks off the deep gravitational well of a rocky planet, it should also be able to remove an Imhotep sized chunk off a weakened slab boundary. What does the flat plane of imhotep tell you, as an EU ideologue?

  • Marco says:

    This sporadic, twice weekly moderation of comments may be to encourage long monologues, rather than the sharing of ideas between interested commenters 🙂 I know that since Philae went to sleep, the mission itself is “between missions” and is looking to be in more or less a low risk mode during the period of solar conjunction. However, now that scientists have actual papers out, and regions have been named, the testing of the logic of various theories by professional and amateur alike are in full swing. The Rosetta blog is the central zone for this citizen science activity and should not just be seen as outreach where the blog is a way for the mission scientists and instrument PI’s to explain new discoveries, or more commonly, to highlight more mysteries. One would hope that the information is going two ways, ie. that any interesting idea that is new and is backed up by new evidence from Rosetta that is explained as a comment on this blog, is indeed being assessed by the scientists concerned where appropriate. A case in point is the monolith(s) on the Amphitheatre as mentioned above. The Amphitheatre is otherwise flat bottomed and roughly circular in shape. This mitigates against the possibility of the monolith being the results of erosion around it . The matching length and size to a nearby ridge on the edge of the Amphitheatre is a strong indication that it broke off the cliff. Lining it up carefully with where it most likely came off reveals a small crater in exactly the right position to straddle both the monolith and the cliff it must have dislodged itself from. This sort of evidence really requires high resolution OSIRIS images to confirm, and it would be very remiss of the science team to completely ignore such fruitful lines of inquiry.

    • emily says:

      Hi Marco,

      There appear to be some misunderstandings, which I shall try to clarify:

      -The blog is moderated in general at least once or twice per working day (depending on workload – remember this outreach team is also working on PR for all other ESA science missions, too!). Some comments will in any case take longer for us to approve if they are feature-length essays 🙂

      -The mission is certainly not “between missions” as you put it; Rosetta is in full-on science phase and will continue to be so throughout the year.

      -The Rosetta blog is here to provide regular updates on the progress of the mission, be it in the form of mission status reports or our regular NAVCAM images, or – when available – scientific reports from the various instrument teams. The comment function is here to facilitate discussions between blog readers and also with us, and is not a platform for any official scientific review process. We very much welcome your discussions and questions.

      -The science teams are of course analysing in detail the data and images they are collecting in order to piece together various lines of enquiry, which obviously includes theories regarding the geological history of the comet. I am sure they are not leaving any stone unturned, so to speak, and we look forward to seeing the results of their analyses in future papers.

      Best wishes,
      Emily and team

      • Marco says:

        Thanks again Emily. It is amazing that even with the Deep Space Network time being taken up more with missions at Ceres and Pluto, together with the difficult sun position, that enough real time data is available both for continuing risk assessment and continuing high resolution science data. I do hope you are right that the mission scientists are leaving no stone unturned. As with everyone else here, we are impatient, but we do understand that the best science takes time. I do see the possibility of non mission scientists getting the jump on new paradigms of comet activity.

  • A.Cooper says:


    Thank you for your continued interest in stretch theory. I have been doing some calculations for the required spin-up rotation speeds for the head lobe to be released from the body. This assumes, of course, that the head lobe was initially sitting on the body without there being any neck at all. It would then have sheared away around the rim of the head and risen above the body. The reasoning behind these calculations then leads on to working out the the slab loss forces and tensile resistance for the Imhotep slab. I’ve explained certain terms for other readers, not for you. 

    The assumed initial shape of the comet would have been a stubby, bulbous pyramid shape with the apex (the current Hatmehit location) slightly offset to one side of centre. Using the dimensions of the comet from the info poster published a few weeks ago, I drew a schematic cross section of this shape and established that the head lobe’s centre of gravity was about 1.8 km from the body lobe’s centre of gravity. 

    I ignored the fact that being so close together their interlocking gravity fields could not strictly allow for the masses to be treated as point sources as they could be if 10 or 20 km apart. But I don’t think this makes more than a 20% difference to the calculations below. 

    For the head lobe to become ‘weightless’ with respect to the body, its c of g would need to be ‘orbiting’ the body c of g at its 1.8 km radius, even though still attached. Feeding this into the orbital velocity equation, v^2= GM(2/r-1/a) gives an orbital speed of 0.603 m/sec. This assumes that the head lobe was integral enough to stay in one lump centered on its c of g- that is a prerequisite assumption of the stretch theory. 

    If it orbited at 0.603 m/sec and was therefore weightless, it would need a greater speed to move away from the body via ‘centrifugal’ force. If it were to escape the body lobe altogether its orbital speed would need to be 0.603 x root 2= 0.863 m/sec due to the balance of kinetic and potential energy. 0.863 m/sec would be its escape velocity. So the rotational speed of the head lobe c of g would be somewhere towards 0.863 m/sec when the head sheared. This speed is the tangential speed at any instantaneous point on the circle it is traveling in around the body c of g. So, seeing as the radius of this circle is 1.8 km its circumference is 11.311 km. Therefore, for the head lobe to be weightless, the rotation period would be 11,311 secs/0.603 metres per sec= 18,758 seconds which is 5 hours and 12 minutes. For it to be at escape velocity, the rotation period would be 11,311/0.863= 13,107 seconds which is 3 hours and 38 minutes. So anywhere between a 3.5-hour to 5-hour rotation period would suffice to lift the head from the body, assuming the tensile resistance was overcome. But a 5-hour rotation would assume virtually no tensile resistance at all.

    Thomas et al (Jan 2015) put an estimate on the tensile strength of the neck itself at 10 pascals and for cliff overhangs, 20 pascals. If we said the initial force required to shear the neck was 40 pascals, it gives a more conservative constraint. 

    So the surplus force (‘centrifugal’) over and above the centripetal force of gravity would need to be 40 pascals and that would be delivered by the difference between mRw^2 for the ‘weightless’ rotation period of 5 hours 12 mins and mRw^2 for some rotation period that gave the 40 pascal surplus. I worked out that if the head was one quarter the mass of the comet and its seating area on the body was 4.68 km squared, the ‘weightless’ value of mRw^2 is 109 pascals. But since it is the weightless scenario, all of that is used up just counteracting the centripetal gravitational force. If we want a 40 pascal surplus, we need 149 pascals. This is 1.37 times more force than the weightless scenario. Since the mass and radius, R remain the same for this new calculation, only the rotation rate, w (omega or radians per second) changes. Since it’s a ‘square’ relationship, w has to increase by the root of 1.37 which means it has to be 1.17 times more than the weightless scenario to give the required 40 pascal surplus. And since w is inversely related to rotation period, the rotation has to be decreased by a factor of 1.17.

     So the rotation period that gives a 40 pascal surplus is 5h12min divided by 1.17 which is 4 hours 26 mins…all based on several assumptions but these are linear assumptions (mass of head lobe, seating area etc) and so are probably within 20 or 30% of the true value, meaning that the 4 hour 26 min value is also within 20-30%. 

    As for the tensile strength of the Imhotep slab, I assumed it was 2000m from the c of g of the whole comet and 100 metres thick, an admittedly highish estimate. It would be weightless at a rotation period of 6 hours 3 mins (calculated using the ‘v squared’ equation and circumference as above). I then calculated the acceleration, Rw^2 for this same radius of 2000m on this rotation period (note, the acceleration, Rw^2 was calculated, not the force, mRw^2 for a reason which will become apparent). This came to 1.66E-4 m/sec^2. This means 0.166 millimeters per second increase in velocity each second. On the other hand, the acceleration, Rw^2 for Imhotep at R=2000 and a 4 hour 26 min period, as suggested for the head lobe shear, is 3.1E-4 m/sec^2. The difference is 1.44E-4m/sec^2 and this would be the surplus acceleration on the Imhotep slab if spun up to 4 hours 26 mins. The actual force per square metre exerted at the Imhotep fracture plane would be mass x acceleration per square metre. The acceleration surplus is 1.44E-4m/sec^2 and the mass is the amount of mass sitting above each square metre of the fracture plane. This is why the mass was left out before and it now gets included. If the slab was 100 metres thick then 100 cubic metres sat above each square metre of the fracture plane. At 470kg/cubic metre, that’s 47,000 kg of mass above each square metre. Therefore the tensile force on each square metre of the imhotep fracture plane for a 100 metre slab and a 4 hour 26 min rotation was 47,000 x 1.44E-4= 6.77 pascals. Not much compared with the head lobe but within the realms of possibility and within 33% of the Thomas et al estimate for the neck. It might seem as though your thoughts on the working free at the fracture plane would need to be invoked.

    However, this low tensile resistance simply can’t be the case if Imhotep was ejected to deep space in one go as is hypothesised. It’s not there any more so it must have left at escape velocity and therefore been doing that velocity prior to release. This in turn means that it must have offered tensile resistance to keep the slab ‘tethered’ as it surpassed its ‘weightless’ or orbital speed for 2000 metres. Otherwise it would have slid sluggishly into an orbit just above the comet. So, to exit at escape velocity it would have to have had the tangential speed of 0.863 m/sec stated above (or thereabouts- the radius is assumed to be 2000m, not 1800m). This corresponds to the 3 hour 38 minute rotation period. Crunching the numbers again gives 13.6 pascals for this shorter period. So maybe the comet spun up to this rotation rate, kept the slab tethered with a 13 pascal or more tensile resistance, and then shed it at escape velocity in one sudden failure episode at around or above 13 pascals. This scenario seems plausible: 3 hr 38 min rotation, 13-15 pascal tethering, escape velocity achieved before tensile failure, slab released to deep space on failure.

    • Robin Sherman says:

      Thank you for that explanation and the ball park calculation, very informative. I am assuming from this that you propose that at some point in the past 67P had a far shorter rotation period than it’s current 12+ hours. Evidence to rule this idea in or out would seem at present to be unobtainable. Some more detailed mechanical properties of the comet material are what is really required.

      As I am posting this reply after you mentioned your post in a later blog and your later ideas about possible lubrication processes, I have to now think that since we are not talking about orders of magnitude differences in energy requirement here, the loss of chunks of comet from the extremities of the lobes due to rotational forces, is looking more and more plausible.

      The large flat surfaces on the ends of the body lobe have always looked to me as if large lumps of the comet have been sheared off. My assumption was that they were probably the result of a glancing blow from an impactor. I struggled with the possibility that more than one of these highly improbable events could have occurred though, there seem to have ben three at least, two on one end and one on the other.

      As I mentioned in my reply to your later post, the migration of denser silicate material towards the top and bottom of the comet, seems to be being born out by the evidence. The extra weight at the ends of the lobes is only going to increase the stress on the neck area. Mass loss at Perihelion has already been accepted as the explanation for the change in rotation period since 2009, this time next year we can expect to see a similar shortening. The predicted loss of up to 20m off the Southern “duckiesphere” surface must amount to a huge mass distribution change. In particular the erosion at the back of the neck is only going to weaken it further. If the head did topple over, would it fall onto the body lobe on top of the Seth region or fall off backwards and become a separate body? Maybe this is a cyclic process even?

      • A.Cooper says:


        I’ve read your reply and your other one on the “Anuket vs Anubis” thread. I thought I’d just recap where we are with ideas on stretch theory so that Marco gets due credit for spin-up, your cryovolcanism is included and I might be able to sell my Roche pass idea a little better than I have done. This recap will hopefully lead seamlessly into answering some if your questions.

        When I suggested the stretch idea last August, Marco replied suggesting spin-up as a candidate mechanism. He had thought of the stretch idea for 67P long before me although I’d been looking at stretching asteroids before my interest in comets commenced on the 6th August 2014. I had suggested the Roche pass at Jupiter as the candidate mechanism: the differential gravitational forces in the pass mimic the effects of spin-up. Marco preferred spin-up because it was a more generic process than chance flybys at Jupiter and would explain the preponderance of peanut-shaped comets. I was very happy to promote Marco’s spin-up ideas here and in my blog series (I’m scute1133) because I know that asteroids spin up to 2 hour rotations via the YORP effect and then shred.

        In August 2014, the density of 470kg/cu metre wasn’t published so I assumed that rotation periods of 4-12 hours would have little bearing on stretch. This was because the assumed density at that time of up to 2000kg/cu metre would mean it would have to spin up to the two-hour asteroid-type speeds or a tiny bit slower. But the very low density was a game changer because it meant a 3 hour 40 min rotation period could achieve escape velocity speeds for the slabs and surplus acceleration for the head lobe. Marco’s spin-up was now all the more believable.

        Your cryovolcanism made it still more plausible. My gas lubrication of pebbles idea was a suggested side-effect of the cryovolcanism and your ‘firn’ phase diagram seemed to back this up. Papers on pebbles and fluffy, deep-fried ice cream dealt a blow to the idea that the comet was a brittle solid lump, impervious to stretch. The density, 70-80% porosity, and 3-metre pebbles were real-life 67P observations that corroborated all of the above. 

        You mentioned whether the forces are enough to separate the lumps of material from the comet and also checked that I did actually mean the comet must have had a shorter rotation period in the past. Yes, I am suggesting that: if the input for the 3h 38 min rotation period is correct (i.e. that the radius of rotation for Imhotep was 1800-2000m when the comet was one solid lump), then that would be enough to shed the end slabs to deep space. As for the mechanism of spin-up, asymmetrical outgassing was cited by Sierks et al (Sept 2014) as causing the 20 min speed up in only five years. Now we have the “20-metre erosion” paper that must also have profound effects on spin-up or spin-down. So I think outgassing and mass redistribution is enough to allow for a 3h 38min period to have occurred by chance in the last few million years or even few thousand years. 

        One intriguing thing is that if the original stubby version of 67P had a period of 3h 38 min, it means that when it stretched due to casing failure, it would have slowed dramatically due to conservation of angular momentum. I haven’t been able to guess at the mass distribution and therefore the changed coefficient of inertia yet. But intuitively I would say that this natural reduction would be a factor of 2 to 3 in rotation period meaning 7h 16 min to 10h 54 min. This is in keeping with the current period and it means we can turn the question of “how did it ever spin up to such a speed when it’s so slow now” to one of “it’s natural speed at one time was 3h 38 min, it sheared under the stress and stretched, hence the much slower rotation period we see today. If we can characterise the coefficient of inertia of the two shapes, the AM conservation may link the two periods near-perfectly and serve as more circumstantial evidence.

        If one objects to outgassing and mass redistribution (and YORP) as sufficient for spin-up, there is always the Roche pass. It requires no spin-up, just relying on delta g forces. However, it is also likely to induce a 4-6 hour rotation anyway but that depends on pre pass rotation characteristics. Given the low density of 67P, this means that this induced rotation accounts for a large portion of the stretch forces during the pass, leaving delta g little extra to do. As I said, the density was unpublished at the time so I wasn’t considering this induced rotation as being significant during the pass. It’s actually highly significant and means that 67P could have passed much higher than the supposed 115,000-130,000 km. That in turn expands the keyhole to perhaps 200,000 km either side of Jupiter or around half a million km. Since Jupiter family comets are thought to have a lifetime between 2.5 million and 32 million years, passing from the centaurs to JFC status an average of 13 times, it means this small keyhole is likely to be found by some if not many JFC’s. Hence it could also be considered as a generic method for stretch, although not as solidly so as for outgassing. It certainly wouldn’t be a vanishingly small chance that a number of JFC’s had gone through the keyhole. The main reason I think 67P is possibly one of these is that its rotation plane is in line with its Jupiter radiant whenever it makes a close pass. The Roche pass would likely align the rotation plane along the radiant. It’s quite a coincidence that it is in line.

        One last point re your question of the head falling onto the body or escaping. I think it would fall onto it. The current potential energy of the head is by definition not equivalent to that required for escape velocity because KE for escape is the same as PE after escape.

        • logan says:

          I imagine 67P kinetics evolving this path: In the short months of this perihelium ‘Coraline’ is going to spin up, significantly. Excepting the occurrence of long duration, tangential main jets.

      • Marco says:

        Hi Andy,
        Thanks for putting together all our thoughts on this, and I agree absolutely with everything that you have said there and earlier. I would just like to add that I believe the lobe masses have self balanced onto the neck due to the neck material being strong under compression and more plastic but weak under tension. Thus i discount the possibility of neck collapse due to erosion or ablation during the southern summer and perihelion. I do expect “Equinox” to be much more interesting than most pundits, as diurnal tidal forces and thermal stresses may combine to create expansion and contraction cycles between the lobes that may induce further stretching of the neck in the cm to metre scale over the equinox period. The imbalance between strength to compressive forces and tensile forces would be one of the factors that would enable this to happen. The visible cracks may expand, or some cryovolcanism could partially seal them from below due to the frictional heat involved. The dust on Hapi may obscure or reveal different cracks as it moves also. There doesn’t seem too much evidence of “remodelling” at least in the Northern hemiduck. Time will tell if the 20metres of erosion on the Southern hemiduck is realistic or if the southern part is similarly static in observable features allowing us to name peaks and spires and boulders before they erode away.

        • A.Cooper says:

          Hi Marco

          I agree that the head won’t collapse any time soon due to the fact that the compression must be good enough to hold up the head. My take on the head dropping but not escaping was more to do with the theoretical scenario of swiping the neck away instantaneously to see what happens. In that case I think it wouldn’t bounce off the body and escape. Perhaps enough neck will be eroded for this to happen one day but I find it very hard to believe these massive 20-metre erosion values. I think there’s something amiss in their model. It’s just a hunch. I haven’t been following the changes as much as others but I do know that you have to look at the same place from many angles to discount white-out effects (high exposure).

Comments are closed.